Platform Services
fix(compliance): correct SYS-ADR-07/Ring-3 imprecision across 4 articles (EN+ES) — 'no runtime call reaches Ring 3' overstated the real rule; corrected to 'Ring 3 never writes, always via proposal + human approval + Ring 2 write path' per architecture/three-ring-architecture.md's own fuller definition. nightly-datagraph-rebuild.md's 'no AI inference/not model-generated' claims were flatly wrong (verified against service-extraction's real code and service-content's Doorman call) — corrected to describe AI-assisted extraction gated by human approval. service-slm-graph-store-migration.md's stale automated-write description corrected with the 2026-07-18 human-approval-gate fix, plus an unrelated short_description/body metadata mismatch (false LadybugDB-to-SQLite migration claim) fixed. Operator-confirmed resolution after joint investigation — not a real compliance breach, just documentation that predated the approval-gate fix.
@@ -16,7 +16,7 @@ paired_with: _index.es.md PointSav's three-ring architecture assigns every service to a layer with defined authority and dependencies. Ring 1 services handle per-tenant boundary ingest — each accepts raw data from one external source and writes it to a durable ledger. Ring 2 services provide deterministic knowledge and processing: they read from Ring 1 and produce structured records, knowledge graphs, and search indexes. Ring 3 is a single service, service-slm, which reads from Ring 2 and never writes to it. The platform functions fully across Rings 1 and 2 without AI compute. Removing Ring 3 shrinks the attack surface, satisfies network-isolation requirements, and answers the compliance question of whether AI has touched the authoritative record. The answer is architectural: Rings 1 and 2 have no import, no dependency, and no runtime call that reaches Ring 3. The platform functions fully across Rings 1 and 2 without AI compute — a deployment can exclude Ring 3 entirely, shrinking the attack surface and satisfying network-isolation requirements. Where Ring 3 is included, the compliance question of whether AI has touched the authoritative record is answered architecturally, not procedurally: Ring 2 services may call Ring 3 for extraction or classification proposals (`service-extraction`'s corpus hand-off to `service-content`, which calls the Doorman for grammar-constrained entity extraction into the DataGraph, is one such path), but Ring 3 never writes to the knowledge graph, the ledger, or any structured record store. Every accepted proposal enters the record only through a Ring 2 write path with a human approval checkpoint. ## Ring 1 — Boundary ingest